February 16, 2021
What’s really happening with OneWeb, Starlink & other Worldwide Low Earth Orbit (LEO) Constellations: As this paper is published, things are moving or “appear” to be moving in the world of low earth orbit internet satellites. In this short piece, we hope to help clear the LEO fog forming in Alaska. When it comes to delivering broadband in Alaska, we don’t necessarily care about who, but more about when, how much, and “are you staying?” STARLINK – When money is no object, you have a large PR staff, and you want to influence regulatory agencies, you start a populist campaign. Starlink is now taking $99 from Alaskans wanting to get in their “better than nothing” test program at some time in 2022. Starlink is primarily focused on direct-to-consumer service in the model of DISH Network and DirecTV. First, let’s discuss a few facts and some educated guesses from what Starlink has publicly released or stated. (1) Starlink does not have legal authority at this point to expand their constellation to serve Alaska. They have applied to the FCC for authority to launch 1240 satellites into orbital inclinations of 70 and 97.6 degrees. They seem to need satellites in both of these inclinations to serve Alaska from Shemya to Utqiaġvik and Ketchikan. Currently they only have temporary authority for the 10 satellites they have currently launched in the 97.6-degree inclination and single gateway for those satellites in Kuparuk. They will need additional gateways in Alaska to serve the state. (2) Starlink launches over the first 18 launches has been 1.2 a month. Each launch normally includes 60 satellites. To complete the orbital inclinations needed for coverage above 50 degrees north latitude will take about 21 more launches. Partial service should be available before then but what partial service means is unclear. (3) Starlink was awarded $880 million from the FCC for providing broadband to unserved households around the United States as part of the Rural Digital Opportunity Fund reverse auction, an auction which excluded households in Alaska. To meet this obligation, Starlink still must launch an additional 2168 satellites in groups of 60 at an inclination around 50 degrees (~36 launches). When will Starlink provide 7x24 service to all of Alaska? It depends on whether they go after the money they can get from the FCC when they are ready to serve the households in the auction or focus on Alaska service. Until they have permission for the 70 degree or 97-degree satellite orbits, any future launches will be for the lower 48 states. At Starlink’s current 1.2 launches per month, it will be 4 years before they are done. Then, they get to start over again as the estimated life of the satellites is 5 years. Starlink’s business model to date is selling direct to customers who self-install their home systems. To serve 40,000 rural Alaska customers will require significant logistic infrastructure. (4) The Starlink user terminal price advertised at $499 is probably highly subsidized and close to the actual OneWeb user terminal price. Estimates have put the Starlink user terminal cost to produce between $1,500 and $2,500 each. ONEWEB – OneWeb is out of bankruptcy and has restarted the gateway in Talkeetna in preparation for initial testing in March with beta testing in April. 24-7 service should be commencing in the fall with the current launch schedule. All of OneWeb’s satellites are in polar orbit and as a benefit, Alaska will be first in the world to get full time service as soon as a little over 200 satellites are in space. OneWeb is primarily an enterprise service, focused on single terminals supporting remote village distribution networks for example. Sites for the initial testing in March have been selected and beta test sites are being solicited by Pacific Dataport. The cost of the OneWeb user terminal does not make it suitable for direct-to-consumer type services, but there will probably be some sales to consumers where a low buy in price is not important. OneWeb has met all legal and regulatory obligations for serving Alaska. They can serve anywhere in Alaska with one single gateway, except for Shemya. OneWeb’s commitment to Alaska must be acknowledged. From their very first business plan, OneWeb made it clear they were going to serve all of Alaska. OneWeb is launching their worldwide service in the state that has the highest amount of unserved and underserved residents. They visited several locations in Alaska and decided to develop the only gateway they’ll need at Microcom’s Talkeetna Alaska Teleport. This central location is perfect and will allow every single OneWeb satellite orbiting over Alaska access to the Internet. Starting this March, OneWeb is launching 36 satellites monthly. Beta testing in Alaska begins in March and by fall, OneWeb will be capable of 24/7 broadband service across Alaska. Preliminary testing showed OneWeb satellites capable of delivering speeds faster than 400 Mbps and average latency of 32ms. It’s worth pointing out that OneWeb is backed by Bharti Global and the UK Government, both fully capable of finishing the OneWeb Network. TELESAT – Telesat recently announced their Lightspeed low earth orbit plans with Thales Alenia Space. Its 298 satellites aimed at worldwide coverage. Telesat is also focusing on the enterprise type market. One interesting point is that any user terminal can communicate directly with any other user terminal making gateways less important. They also intend to make extensive use of optical (read laser) links between satellites so that user terminal can communicate with any other user terminal in the world without accessing a gateway. It is so important to them that they recently received a $30 million grant from the Defense Advanced Research Projects Agency to develop the capability. It would seem to indicate they need more R&D to assure what they plan to do will work. More important is that Telesat still has not raised the capital to build the constellation of satellites. They have a commitment from the government of Canada for about 20% of what they need but that still leaves a long way to go. KUIPER (Amazon) – We can’t say much here as so little is known about their plans or timeline. Like Starlink, money is not an object with this organization. Amazon may be Kuiper’s best customer as they begin providing services to Amazon Web Services (AWS). LEO CHALLENGES IN ALASKA We’ll close our little Alaska review of low earth orbit satellite plans with some comments on problems common to all these operators and their plans. (1) Money. All of these networks require a lot of money before one dollar is earned, $3-12 billion by many estimates. With 60% of the earth’s surface representing oceans, then 60% of the earth doesn’t represent much of a revenue stream except for the occasional airplane or ship. Look at those parts of the earth that have limited ability to pay even $99 a month, and you are left with North America, Europe, parts of Asia (Russia and China are particularly distrustful of systems owned and operated by western powers), and a few other scattered locations. That makes for an interesting business plan. You’d have to share the frequencies you are allowed to use with 3 other operators (4 if you count O3B/M-Power operating between 50N and 50S) and negotiate agreements with 195 countries and hope you are not locked out of the big ones by a competitor or politics. Bankruptcy seems to be a rite of passage for LEO’s, as most LEO service providers to date have faced bankruptcy. (2) These businesses must succeed worldwide and not just Alaska. That is a difficult journey as each country levies different requirements on the business for permission to operate in the country. For true worldwide service you need 195 agreements impacting how the internet data is routed, what it costs, how the common frequencies used by all operators are shared (more on that), and who even gets to operate in the country. (3) All of these networks operate in the electromagnetic spectrum (radio for the uninitiated). There are specific parts of that spectrum (frequencies) allocated to low earth orbit operations internationally by the International Telecommunications Union. With multiple companies wanting to do the same thing, there has to be some framework for allocating the frequencies. That’s where it gets complicated. The ITU uses a first come first served approach meaning the first one to apply for use of frequencies gets priority for use. Everyone else is behind them in the line. However, this in only true over international waters and Antarctica. Any operator that is claiming priority rights to spectrum (OneWeb and Telesat) are only speaking in terms of international waters and Antarctica. Each country can adopt the ITU policy or set their own and the US has decided that the frequencies must be shared equally. That means if two operators want to use Ku band frequencies, each gets to use half. In the U.S., all four networks above; Starlink, OneWeb, Telesat, and Kuiper will be subject to the frequency sharing regime. That’s a big deal no one is talking about or understands how it will work. (4) It is estimated by industry executives that only two or three “free market” competing LEO’s will actually survive full market penetration. In other words, based on the number of potential customers and revenue, there’s only enough business for two or three LEO’s. First mover advantage, innovative discoveries, and large bank accounts will help determine which LEO’s survive. This does not include state-sponsored LEO systems. https://docs.fcc.gov/public/attachments/DA-21-34A1.pdf https://en.wikipedia.org/wiki/Starlink https://spacenews.com/spacex-wins-big-share-of-9-2b-rdof-broadband-subsidy/ https://spacenews.com/oneweb-emerges-from-chapter-11-with-new-ceo/ https://www.akbizmag.com/industry/telecom-tech/pacific-dataport-preps-for-oneweb-beta-testing-in-alaska/ https://www.alaskateleport.com/ https://www.telesat.com/press/press-releases/manufacturer-announcement/ https://www.telesat.com/press/press-releases/telesat-u-s-services-awarded-darpa-contract-for-blackjack-track-b-research-development-and-in-orbit-demonstration-with-telesat-leo/ https://www.telesat.com/press/press-releases/telesat-and-the-government-of-canada-finalize-600m-agreement-to-bridge-canadas-digital-divide-with-telesats-low-earth-orbit-satellite-constellation/ https://en.wikipedia.org/wiki/Kuiper_Systems https://www.itu.int/en/ITU-R/space/workshops/2019-SatSymp/Presentations/106%20-%20NGSO%20Large%20Constellations%20FCC-USA.pdf